Download attribute feature is to download a content instead of navigating it. For enabling this, the FrameLoaderClientEfl::startDownload should get a suggestedName, too. Additionally, for layout tests, a callback for "download,request" is needed on DumpRenderTree for EFL.
Created attachment 163508 [details] Patch
Comment on attachment 163508 [details] Patch View in context: https://bugs.webkit.org/attachment.cgi?id=163508&action=review > Tools/DumpRenderTree/efl/DumpRenderTreeChrome.cpp:897 > + Evas_Object* newView = browser->createView(); IMO, it would be good if you say why this callback needs to create new view via a comment.
Comment on attachment 163508 [details] Patch View in context: https://bugs.webkit.org/attachment.cgi?id=163508&action=review > LayoutTests/platform/efl/Skipped:-154 > -fast/dom/HTMLAnchorElement/anchor-download-unset.html Please check that those tests pass for WebKit2 EFL. If they don't, move them to LayoutTests/efl-wk2/TestExpectations. We don't want to make the WK2 bot red.
Created attachment 163571 [details] Patch
Comment on attachment 163571 [details] Patch View in context: https://bugs.webkit.org/attachment.cgi?id=163571&action=review > Source/cmake/OptionsEfl.cmake:57 > +WEBKIT_OPTION_DEFAULT_PORT_VALUE(ENABLE_DOWNLOAD_ATTRIBUTE ON) Wrong alphabetical order.
Created attachment 163588 [details] Patch
Comment on attachment 163588 [details] Patch View in context: https://bugs.webkit.org/attachment.cgi?id=163588&action=review > LayoutTests/platform/efl-wk2/TestExpectations:199 > +BUGWKEFL : fast/dom/HTMLAnchorElement/anchor-download-unset.html = TEXT PASS Passes for me on WK2, did you check? > LayoutTests/platform/efl-wk2/TestExpectations:200 > +BUGWKEFL : fast/dom/HTMLAnchorElement/anchor-download.html = TEXT PASS Fails on WK2 but displays "PASS". Needs to be investigated to have a proper comment in TestExpectations. > LayoutTests/platform/efl-wk2/TestExpectations:201 > +BUGWKEFL : fast/dom/HTMLAnchorElement/anchor-nodownload-set.html = TEXT PASS Ditto. > LayoutTests/platform/efl-wk2/TestExpectations:202 > +BUGWKEFL : fast/dom/HTMLAnchorElement/anchor-nodownload.html = TEXT PASS Passes for me on WK2.
Created attachment 163593 [details] Patch
The test cases anchor-download-unset.html and anchor-nodownload.html is passed although the download_attribute is not supported. But what I intended was the 4 test cases should be checked althogether (I mentioned it on a comment). As your comment, because that can be confused, I modified to add only failed test cases on a TestExpectation file.
Additionally, I'll explain the test cases below. anchor-download.html is for an anchor tag that has 'download' as attribute - <a download> anchor-download-unset.html is a test for removing the download attribute by scripts. anchor-nodownload-set.html is for adding the download attribute to the anchor tag that have no download attribute by scripts. anchor-nodownload.html is for anchor tag that does not have download attribute. So, 'anchor-download.html' and 'anchor-nodownload-set.html' should have same result, and 'anchor-download-unset.html' and 'anchor-nodownload.html' should have same result. In the test cases, the blob url that the anchor tag linked is a new document that has "PASS" text as contents. If there is no download attribute on the anchor tag ('anchor-download-unset.html' and 'anchor-nodownload.html'), when the anchor tag is clicked, the current view should load the blob url, that has "PASS" text as contents. If there is download attribute on the anchor tag ('anchor-download.html' and 'anchor-nodownload-set.html'), when the anchor tag is clicked, the blob url should not be loaded on a current view (should be downloaded), so the current contents "Blob URL" should be remained. So, if the test case 'anchor-download.html' has "PASS" text as contents, that means it was failed.
Comment on attachment 163593 [details] Patch View in context: https://bugs.webkit.org/attachment.cgi?id=163593&action=review > Tools/DumpRenderTree/efl/DumpRenderTreeChrome.cpp:900 > + Evas_Object* newView = browser->createView(); Nit : wrong indentation.
Created attachment 163734 [details] Patch
Comment on attachment 163734 [details] Patch Looks fine now.
Comment on attachment 163734 [details] Patch Clearing flags on attachment: 163734 Committed r128393: <http://trac.webkit.org/changeset/128393>
All reviewed patches have been landed. Closing bug.